Financial Crimes Compliance Modeling & Analytics Manager

<p>Mercury is building a complete finance stack for startups. We work hard to create the easiest and safest banking* experience possible to simplify entrepreneurs' and business owners' financial lives. The challenge is to do so while ensuring we protect Mercury, customers, and the broader financial ecosystem from bad actors and harmful, illegal, or unauthorized activities.</p> <p>Mercury builds banking* for ambitious entrepreneurs. While we’re not a bank ourselves, our work sits within the broader U.S. financial system — and with that comes a responsibility to help keep it safe. That means building thoughtful systems and processes that root out bad actors, prevent abuse, and protect access for the founders and companies we’re here to serve.</p> <p>The BSA/AML &amp; Sanctions compliance team serves as the oversight function for Mercury's overall AML &amp; Sanctions program. As Financial Crimes Compliance Modeling &amp; Analytics Manager, you'll help drive enhancements to Mercury's financial crimes compliance (FCC) detection and screening models and improve the overall FCC framework. You'll play a key role in developing, tuning, and maintaining Mercury's transaction monitoring (TM) and Sanctions models, and in building the analytics and metrics that track the health of FCC programs. This technical role requires a strong analytical background and FCC subject-matter expertise, and will collaborate closely with risk strategy, engineering, and compliance teams.</p> <p><em>*</em><em>Mercury is a fintech company, not an FDIC-insured bank. Banking services provided through Choice Financial Group and Column N.A., Members FDIC</em>.</p> <p>&nbsp;</p> <h3>As part of the journey, we would expect you to:</h3> <ul> <li>Use SQL and other analytical tools to conduct in-depth analysis of Mercury's customers, transactions, alerts, TM rules, risk ratings, and more</li> <li>Use data-driven methods to improve, design, implement, and maintain Mercury's FCC models, including transaction monitoring, sanctions screening, and relevant models</li> <li>Develop bespoke transaction monitoring rules and sanctions screening logic designed to address Mercury's specific AML and sanctions risk</li> <li>Partner with Compliance, Product, and Data leaders to translate regulatory requirements into effective analytical frameworks</li> <li>Know how to tell stories with data, enabling people to understand the output and meaning of analytics activities in a clear, compelling manner</li> <li>Interpret analytics outputs to pinpoint which alerts, patterns, or anomalies signal genuine risk, and articulate why they matter to compliance and business stakeholders</li> <li>Develop and maintain detailed documentation on the configuration of FCC models including scenarios, thresholds, segments, tuning, false positive rules, etc., and any changes made to those configurations over time</li> <li>Evaluate and tune existing detection models and rules to reduce false positives while maintaining regulatory rigor</li> <li>Develop data-driven methods to identify new typologies, emerging risks, and evolving financial crime trends</li> <li>Partner with Model Risk Management to support validation and performance monitoring of models to ensure compliance with internal and regulatory standards</li> </ul> <h3>There are lots of paths that could lead you to be successful in a role like this; we think the strongest candidates will have some combination of the following:</h3> <ul> <li>Bachelor's degree in a quantitative field (e.g. Computer Science, Engineering, Statistics, Mathematics, or related) with 8+ years of experience conducting in-depth data analytics, ideally with 5+ years in FCC or AML/Sanctions related analytics roles</li> <li>Deep understanding of AML and Sanctions fundamentals, including both principles and regulations</li> <li>Outstanding skills with standard analytical tools; top-notch SQL skills required, experience with Python or similar preferred, and familiarity with modern ML tooling (e.g. scikit-learn, XGBoost) a plus</li> <li>Experience developing, tuning, and maintaining machine learning or rule-based detection models, with an understanding of how to rigorously challenge model performance and limitations</li> <li>Experience identifying ways to improve both data-related and operational efficiencies</li> <li>A healthy dose of skepticism combined with a constructive, solution-oriented approach</li> <li>Comfort operating with ambiguity and capable of synthesizing fragmented technical, operational, and business context into a clear understanding of how models actually work, even with